With the development of computer communication technologies and popularity of the Internet, information interaction via instant messaging clients has become one of important contact means among people nowadays. However, interacting parties generally receive only texts or multimedia information such as audio from the counterpart but cannot see appearance of the counterpart, thus malefactors can simulate a legal user identity of a stolen instant messaging account to perform a money fraud to a contact of the instant messaging account. Therefore, it is an essential subject to prompt a user to guarantee information safety in the process of instant messaging.
Nowadays, safety pre-warnings in instant messaging processes generally cannot accurately identify dangerous situations, resulting in low accuracy of the safety pre-warnings.